INDEX

Explanations

references to comprehensiveness and inclusivity in various contexts

oai_token-act-pair · gpt-4o-mini Triggered by @bot

New Auto-Interp

Top Features by Cosine Similarity

Comparing With LLAMA3.1-8B @ 26-llamascope-res-32k

Configuration

fnlp/Llama3_1-8B-Base-LXR-8x/Llama3_1-8B-Base-L26R-8x

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

cerebras/SlimPajama-627B

Features

32,768

Data Type

bfloat16

Hook Name

blocks.26.hook_resid_post

Hook Layer

Architecture

jumprelu

Context Size

1,024

Dataset

cerebras/SlimPajama-627B

Activation Function

relu

Embeds

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

/commons

-0.15

.gdx

-0.14

aln

-0.14

 EntityState

-0.13

å¢

-0.13

osti

-0.13

 ç·¨

-0.13

/socket

-0.13

ursively

-0.13

tractor

-0.13

POSITIVE LOGITS

 aspects

0.94

 aspect

0.94

aspect

0.82

 Aspect

0.75

Aspect

0.69

 facets

0.68

 facet

0.59

æĸ¹éĿ¢

0.53

pects

0.52

 Ð°ÑģÐ¿

0.52

Activations Density 0.205%

references to comprehensiveness and inclusivity in various contexts

No Comments

No Known Activations